Panoptic Feature Pyramid Network is an end-to-end panoptic segmentation network proposed by Facebook AI Research that uses Mask R-CNN for semantic segmentation tasks. It attaches a newly proposed semantic branch to Mask R-CNN to perform pixel-level classification for stuff classes. Experiments showed it achieved competitive results on standard datasets with less memory usage compared to other single panoptic segmentation networks. UPSNet is another unified panoptic segmentation network from Uber that uses Mask R-CNN with semantic and panoptic heads to merge outputs for panoptic segmentation. It also achieved state-of-the-art performance on benchmarks while using fewer parameters.